Measurement instruments are used to test and analyze signals from electrical and electronic devices and systems. Examples of such measurement instruments include spectrum analyzers and vector signal analyzers which may be employed to characterize the frequency and/or phase responses of signals in the RF, microwave and/or millimeter wave frequency spectra.
Calibration is an important characteristic of such measurement instruments. That is, when using such a measurement instrument to make measure a characteristic (e.g., frequency response) of a signal (e.g., an output signal of a device under test (DUT)), it is important to know the amplitude and phase response of the measurement channel of the measurement instrument. In some cases, it may be sufficient to know that the measurement channel is operating within the specified parameters of the measurement instrument—for example that the amplitude and phase response as a function of frequency are within certain specified ranges or tolerances across a specified frequency span or range. In other cases, it may be desired to know the current operating parameters of the measurement instrument—for example the amplitude and phase response as a function of frequency across a specified frequency span or range—so that any deviations from an ideal characteristic in the measurement channel may be compensated when making signal measurements.
In particular, several techniques have been employed for characterizing the amplitude and phase linearity as a function of frequency of a signal processing channel of a wide bandwidth spectrum analyzer operating in the zero-span, digitizing (non-swept) mode.
In one known technique, a baseband square-wave reference signal is routed to the RF input of the analyzer, the analyzer is tuned to 0 Hz center frequency, and the in-phase/quadrature-phase (I/Q) waveform is captured by an intermediate frequency (IF) digitizer internal to the analyzer. By post-processing the I/Q record, the phase and amplitude channel response of the analyzer's IF channel can be measured and, if desired, used for flatness compensation. However, there are some disadvantages or limitations to this technique. First, the technique only makes measurements with the analyzer set to its low-band IF path. However, the analyzer may have multiple IF paths. So this technique can only measure the lowest-frequency heterodyne mixing band of the analyzer. Second, the measurement captures only phase and amplitude frequency response at a single RF frequency (0 Hz).
Another known technique involves routing a carrier wave (CW) signal to the RF input of the analyzer, tuning the analyzer to the CW center frequency, and sweeping the analyzer's local oscillator (LO) linearly across a frequency span of interest. The sweep of the LO signal causes the CW signal in turn to be swept across the IF bandwidth, and this sweep may be captured by the analyzer's IF digitizer, and the resulting I/Q record can be post-processed to calculate the phase and amplitude channel response of the analyzer's IF channel. However, there are also some disadvantages or limitations to this technique. First, this technique requires exact time alignment between the LO sweep and digitizer I/Q capture. This alignment is also subject to phase errors due to any non-linearity of the LO sweep. Also, in the normal “use” of the analyzer making a wideband frequency measurement of a signal (e.g., an output signal of a DUT), the LO is at a fixed frequency while the digitizer captures the IF signal across the IF channel. However since the LO is not fixed but is swept during the channel measurement procedure, frequency response errors are introduced into the channel measurement due to variations in the LO frequency.
Still another known technique employs an impulse generator, such as the KEYSIGHT® U9391 Comb Generator which may be driven by an RF reference signal (typically 10 MHz) to create a comb signal having a spectrum of equally-spaced tones with known phase and amplitude characteristics. This comb signal routed to the RF input of the analyzer, the analyzer is tuned to the desired center frequency in zero-span mode, and the I/Q waveform is captured by an internal IF digitizer of the analyzer. By post-processing the I/Q record, the phase and amplitude channel response of the analyzer's IF can be measured. However, there are also some disadvantages or limitations to this technique. First, as the analyzer center frequency increases, the power levels of the comb harmonics decrease greatly. The low signal-to-noise ratio of the measured signal degrades the repeatability of the measurement. Second, signal amplification to increase the power level of the comb signal's higher frequency harmonics causes the comb generator to consume more power. For example, the U9391G Comb Generator uses 13 watts to generate −85 dBm tones at 67 GHz.
It would be desirable to provide another technique for characterizing the amplitude and phase linearity as a function of frequency of a signal processing channel, for example a signal processing channel (e.g., an IF channel) of a spectrum analyzer or vector signal analyzer which may avoid some or all of the disadvantages or limitations of existing techniques.